Peer Review, Experiments & Background Information

Peer review (known as refereeing in some academic fields) is a process of subjecting an author's scholarly work, research or ideas to the scrutiny of others who are experts in the same field. It is used primarily by editors to select and to screen submitted manuscripts, and by funding agencies to decide the awarding of grants. The peer review process has a normative function by encouraging authors to meet the accepted high standards of their discipline and to prevent the dissemination of unwarranted claims, unacceptable interpretations or personal views. Publications and awards that have not undergone peer review are likely to be regarded with suspicion by scholars and professionals in many fields. Even refereed journals, however, can contain errors.

In the case of manuscripts, the editor will pass manuscripts that are accepted for publication to a publisher who will be responsible for organizing redactory services, printing and distribution of the publication. In specialist academic (scholarly) journals, the editor (or increasingly group of editors) is normally a well-respected academic in the field, and edits the journal on behalf of a learned society or a commercial publisher. Some journals have professional editors employed by the owner of the journal. An editor is ultimately responsible for the quality and selection of manuscripts chosen to be published, usually basing their decision on peer review, although the authors are always responsible for the content of each manuscript. The editor does not revise and correct spelling, grammar and formatting — that process is carried out by a copy editor, although the editor controls the quality of the process.

Reasons for peer review

A common rationale for peer review is that it is rare for an individual author or research team to spot every mistake or flaw in a complicated piece of work. This is not because deficiencies represent "needles in a haystack" that are difficult to locate, but because with a new and perhaps eclectic subject, an opportunity for improvement may be more obvious to someone with special expertise or experience. For both grant-funding and publication in a scholarly journal, it is also normally a requirement that the subject is both novel and substantial. Therefore, showing work to others increases the probability that weaknesses will be identified, and, with advice and encouragement, fixed.

Reviewers are typically anonymous and independent, to help foster unvarnished criticism, and to discourage cronyism in funding and publication decisions. However, as discussed below under the next section, US government guidelines governing peer review for federal regulatory agencies require that reviewer's identity be disclosed under some circumstances.

In addition, since reviewers are normally selected from experts in the fields discussed in the article, the process of peer review is considered critical to establishing a reliable body of research and knowledge. Scholars reading the published articles can only be expert in a limited area; they rely, to some degree, on the peer-review process to provide reliable and credible research that they can build upon for subsequent or related research. As a result, significant scandal ensues when an author is found to have falsified the research included in an article, as many other scholars, and the field of study itself, may have relied upon the original research (see Peer review and fraud below).

How it works

In the case of proposed publications, an editor sends advance copies of an author's work or ideas to researchers or scholars who are experts in the field (known as "referees" or "reviewers"), normally by e-mail or through a web-based manuscript processing system. Usually, there are two or three referees for a given article.

These referees each return an evaluation of the work to the editor, including noting weaknesses or problems along with suggestions for improvement. Typically, most of the referees' comments are eventually seen by the author; scientific journals observe this convention universally. The editor, usually familiar with the field of the manuscript (although typically not in as much depth as the referees, who are specialists), then evaluates the referees' comments, her or his own opinion of the manuscript, and the context of the scope of the journal or level of the book and readership, before passing a decision back to the author(s), usually with the referees' comments.

Referees' evaluations usually include an explicit recommendation of what to do with the manuscript or proposal, often chosen from a menu provided by the journal or funding agency. Most recommendations are along the lines of the following:

During this process, the role of the referees is advisory, and the editor is typically under no formal obligation to accept the opinions of the referees. Furthermore, in scientific publication, the referees do not act as a group, do not communicate with each other, and typically are not aware of each other's identities or evaluations. There is usually no requirement that the referees achieve consensus. Thus the group dynamics are substantially different from that of a jury. In situations where the referees disagree substantially about the quality of a work, there are a number of strategies for reaching a decision.

When an editor receives very positive and very negative reviews for the same manuscript, the editor often will solicit one or more additional reviews as a tie-breaker. As another strategy in the case of ties, editors may invite authors to reply to a referee's criticisms and permit a compelling rebuttal to break the tie. If an editor does not feel confident to weigh the persuasiveness of a rebuttal, the editor may solicit a response from the referee who made the original criticism. In rare instances, an editor will convey communications back and forth between authors and a referee, in effect allowing them to debate a point. Even in these cases, however, editors do not allow referees to confer with each other, and the goal of the process is explicitly not to reach consensus or to convince anyone to change their opinions. Some medical journals, however (usually following the open access model), have begun posting on the Internet the pre-publication history of each individual article, from the original submission to reviewers' reports, authors' comments, and revised manuscripts.

Traditionally, reviewers would remain anonymous to the authors, but this standard is slowly changing. In some academic fields, most journals now offer the reviewer the option of remaining anonymous or not, or a referee may opt to sign a review, thereby relinquishing anonymity. Published papers sometimes contain, in the acknowledgements section, thanks to anonymous or named referees who helped improve the paper.

Some university presses undertake peer review of books. After positive review by two or three independent referees, a university press sends the manuscript to the press's editorial board, a committee of faculty members, for final approval.^[1] Such a review process is a requirement for full membership of the Association of American University Presses.^[2]

Recruiting referees

At a journal or book publisher, the task of picking reviewers typically falls to an editor.^[3] When a manuscript arrives, an editor solicits reviews from scholars or other experts who may or may not have already expressed a willingness to referee for that journal or book division. Granting agencies typically recruit a panel or committee of reviewers in advance of the arrival of applications.

In some disciplines there exist refereed venues (such as conferences and workshops). To be admitted to speak, scholars and scientists must submit papers (generally short, often 15 pages or less) in advance. These papers are reviewed by a "program committee" (the equivalent of an editorial board), which generally requests inputs from referees. The hard deadlines set by the conferences tend to limit the options to either accept or reject the paper.

Typically referees are not selected from among the authors' close colleagues, students, or friends. Referees are supposed to inform the editor of any conflict of interests that might arise. Journals or individual editors often invite a manuscript's authors to name people whom they consider qualified to referee their work. Authors are sometimes also invited to name natural candidates who should be disqualified, in which case they may be asked to provide justification (typically expressed in terms of conflict of interest). In some disciplines, scholars listed in an "acknowledgements" section are not allowed to serve as referees (hence the occasional practice of using this section to disqualify potentially negative reviewers).

Editors solicit author input in selecting referees because academic writing typically is very specialized. Editors often oversee many specialities, and may not be experts in any of them, since editors may be full time professionals with no time for scholarship. But after an editor selects referees from the pool of candidates, the editor typically is obliged not to disclose the referees' identities to the authors, and in scientific journals, to each other. Policies on such matters differ among academic disciplines.

Recruiting referees is a political art, because referees, and often editors, are usually not paid, and reviewing takes time away from the referee's main activities, such as his or her own research. To the would-be recruiter's advantage, most potential referees are authors themselves, or at least readers, who know that the publication system requires that experts donate their time. Referees also have the opportunity to prevent work that does not meet the standards of the field from being published, which is a position of some responsibility. Editors are at a special advantage in recruiting a scholar when they have overseen the publication of his or her work, or if the scholar is one who hopes to submit manuscripts to that editor's publication in the future. Granting agencies, similarly, tend to seek referees among their present or former grantees. Serving as a referee can even be a condition of a grant, or professional association membership.

Another difficulty that peer-review organizers face is that, with respect to some manuscripts or proposals, there may be few scholars who truly qualify as experts. Such a circumstance often frustrates the goals of reviewer anonymity and the avoidance of conflicts of interest. It also increases the chances that an organizer will not be able to recruit true experts – people who have themselves done work similar to that under review, and who can read between the lines. Low-prestige or local journals and granting agencies that award little money are especially handicapped with regard to recruiting experts.

Finally, anonymity adds to the difficulty in finding reviewers in another way. In scientific circles, credentials and reputation are important, and while being a referee for a prestigious journal is considered an honor, the anonymity restrictions make it impossible to publicly state that one was a referee for a particular article. However, credentials and reputation are principally established by publications, not by refereeing; and in some fields refereeing may not be anonymous.

The process of peer review does not end after a paper completes the peer review process. After being put to press, and after 'the ink is dry', the process of peer review continues in journal clubs. Here groups of colleagues review literature and discuss the value and implications it presents. Journal clubs will often send letters to the editor of a journal, or correspond with the editor via an on-line journal club. In this way, all 'peers' may offer review and critique of published literature.

Different styles of review

Peer review can be rigorous, in terms of the skill brought to bear, without being highly stringent. An agency may be flush with money to give away, for example, or a journal may have few impressive manuscripts to choose from, so there may be little incentive for selection. Conversely, when either funds or publication space is limited, peer review may be used to select an extremely small number of proposals or manuscripts.

Often the decision of what counts as "good enough" falls entirely to the editor or organizer of the review. In other cases, referees will each be asked to make the call, with only general guidance from the coordinator on what stringency to apply.

Very general journals such as Science and Nature have extremely stringent standards for publication, and will reject papers that report good quality scientific work if editors feel the work is not a breakthrough in the field. Such journals generally have a two-tier reviewing system. In the first stage, members of the editorial board verify that the paper's findings — if correct — would be ground-breaking enough to warrant publication in Science or Nature. Most papers are rejected at this stage. Papers that do pass this 'pre-reviewing' are sent out for in-depth review to outside referees. Even after all reviewers recommend publication and all reviewer criticisms/suggestions for changes have been met, papers may still be returned to the authors for shortening to meet the journal's length limits. With the advent of electronic journal editions, overflow material may be stored in the journal's online Electronic Supporting Information archive.

A similar emphasis on novelty exists in general area journals such as the Journal of the American Chemical Society (JACS). However, these journals generally send out all papers (except blatantly inappropriate ones) for peer reviewing to multiple reviewers. The reviewers are specifically queried not just on the scientific quality and correctness, but also on whether the findings are of interest to the general area readership (chemists of all disciplines, in the case of JACS) or only to a specialist subgroup. In the latter case, the recommendation is usually for publication in a more specialized journal. The editor may offer to authors the option of having the manuscript and reviews forwarded to such a journal with the same publishers (e.g., in the example given, Journal of Organic Chemistry, Journal of Physical Chemistry, Inorganic Chemistry,...) if the reviewer reports warrant such a decision (i.e., they boil down to "Great work, but too specialized for JACS: publish in ..."), the editor of such a journal may accept the forwarded manuscript without further reviewing.

Some general area journals, such as Physical Review Letters, have strict length limitations. Others, such as JACS, have Letters and Full Papers sections: the Letters sections have strict length limits (two journal pages in the case of JACS) and special novelty requirements. In contrast, online-only journals may have no space limitations.^[4] More specialized scientific journals such as the aforementioned chemistry journals, Astrophysical Journal, and the Physical Review series use peer review primarily to filter out obvious mistakes and incompetence, as well as plagiarism, overly derivative work, and straightforward applications of known methods. Different publication rates reflect these different criteria: Nature publishes about 5 percent of received papers, while Astrophysical Journal publishes about 70 percent. The different publication rates are also reflected in the size of the journals. PLoS ONE was launched by the Public Library of Science in 2006 with the aim to "concentrate on technical rather than subjective concerns", and to publish articles from across science, regardless of the field.^[5]. Another open access journal, Biology Direct, has the policy of making the reviewers' reports public by publishing the reports together with the manuscripts.

Screening by peers may be more or less laissez-faire depending on the discipline. Physicists, for example, tend to think that decisions about the worthiness of an article are best left to the marketplace. Yet even within such a culture peer review serves to ensure high standards in what is published. Outright errors are detected and authors receive both edits and suggestions.

To preserve the integrity of the peer-review process, submitting authors may not be informed of who reviews their papers; sometimes, they might not even know the identity of the associate editor who is responsible for the paper. In many cases, alternatively called "masked" or "double-masked" review (or "blind" or "double-blind" review), the identity of the authors is concealed from the reviewers, lest the knowledge of authorship bias their review; in such cases, however, the associate editor responsible for the paper does know who the author is. Sometimes the scenario where the reviewers do know who the authors are is called "single-masked" to distinguish it from the "double-masked" process. In double-masked review, the authors are required to remove any reference that may point to them as the authors of the paper.

While the anonymity of reviewers is almost universally preserved, double-masked review (where authors are also anonymous to reviewers) is still relatively rarely employed.

Critics of the double-masked process point out that, despite the extra editorial effort to ensure anonymity, the process often fails to do so, since certain approaches, methods, writing styles, notations, etc., may point to a certain group of people in a research stream, and even to a particular person.^[6], ^[7] Proponents of double-masked review argue that it performs at least as well as the traditional one and that it generates a better perception of fairness and equality in global scientific funding and publishing.^[8]

Proponents of the double-masked process argue that if the reviewers of a paper are unknown to each other, the associate editor responsible for the paper can easily verify the objectivity of the reviews. Single-masked review is thus strongly dependent upon the goodwill of the participants.

A more rigorous standard of accountability is known as an audit. Because reviewers are not paid, they cannot be expected to put as much time and effort into a review as an audit requires. Most journals (and grant agencies like NSF) have a policy that authors must archive their data and methods in the event another researcher wishes to replicate or audit the research after publication. Unfortunately, the archiving policies are sometimes ignored by researchers.

Criticisms of peer review

One of the most common complaints about the peer review process is that it is slow, and that it typically takes several months or even several years in some fields for a submitted paper to appear in print. In practice, much of the communication about new results in some fields such as astronomy no longer takes place through peer reviewed papers, but rather through preprints submitted onto electronic servers such as arXiv.org. However, such preprints are often also submitted to refereed journals, and in many cases have, at the time of electronic submission, already passed through the peer review process and been accepted for publication.

While passing the peer-review process is often considered in the scientific community to be a certification of validity, it is not without its problems. Drummond Rennie, deputy editor of Journal of the American Medical Association is an organizer of the International Congress on Peer Review and Biomedical Publication, which has been held every four years since 1986.^[9] He remarks, "There seems to be no study too fragmented, no hypothesis too trivial, no literature too biased or too egotistical, no design too warped, no methodology too bungled, no presentation of results too inaccurate, too obscure, and too contradictory, no analysis too self-serving, no argument too circular, no conclusions too trifling or too unjustified, and no grammar and syntax too offensive for a paper to end up in print."^[10]

Richard Horton, editor of the British medical journal The Lancet, has said that "The mistake, of course, is to have thought that peer review was any more than a crude means of discovering the acceptability — not the validity — of a new finding. Editors and scientists alike insist on the pivotal importance of peer review. We portray peer review to the public as a quasi-sacred process that helps to make science our most objective truth teller. But we know that the system of peer review is biased, unjust, unaccountable, incomplete, easily fixed, often insulting, usually ignorant, occasionally foolish, and frequently wrong." ^[11]

Allegations of bias and suppression

In addition, some sociologists of science argue that peer review makes the ability to publish susceptible to control by elites and to personal jealousy.^[12] The peer review process may suppress dissent against "mainstream" theories.^[13]^[14]^[15] Reviewers tend to be especially critical of conclusions that contradict their own views, and lenient towards those that accord with them. At the same time, elite scientists are more likely than less established ones to be sought out as referees, particularly by high-prestige journals or publishers. As a result, it has been argued, ideas that harmonize with the elite's are more likely to see print and to appear in premier journals than are iconoclastic or revolutionary ones, which accords with Thomas Kuhn's well-known observations regarding scientific revolutions.^[16]

Others have pointed out that there is a very large number of scientific journals in which one can publish, making total control of information difficult. In addition, the decision-making process of peer review, in which each referee gives their opinion separately and without consultation with the other referees, is intended to mitigate some of these problems. Some have suggested that:

Nonetheless, while it is generally possible to publish results somewhere, in order for scientists in many fields to attract and maintain funding it is necessary to publish in elite, prestigious journals. Such journals are generally identified by their impact factor. The small number of high-impact journals is susceptible to control by an elite group of anonymous reviewers. Results published in low-impact journals are usually ignored by most scientists in any field. This has led to calls for the removal of reviewer anonymity (especially at high-impact journals) and for the introduction of author anonymity (so that reviewers cannot tell whether the author is a member of any elite).

Peer review failures

Peer review failures occur when a peer-reviewed article contains obvious fundamental errors that undermines at least one of its main conclusions. Peer review is not considered a failure in cases of deliberate fraud by authors. Letters-to-the-editor that correct major errors in articles are a common indication of peer review failures. Many journals have no procedure to deal with peer review failures beyond publishing letters.^[18] Some do not even publish letters. The author of a disputed article is allowed a published reply to a critical letter. Neither the letter nor the reply is usually peer-reviewed, and typically the author rebuts the criticisms. Thus, the readers are left to decide for themselves if there was a peer review failure.

Peer review, in scientific journals, assumes that the article reviewed has been honestly written, and the process is not designed to detect fraud. The reviewers usually do not have full access to the data from which the paper has been written and some elements have to be taken on trust. It is not usually practical for the reviewer to reproduce the author's work, unless the paper deals with purely theoretical problems which the reviewer can follow in a step-by-step manner.

The number and proportion of articles which are detected as fraudulent at review stage is unknown. Some instances of outright scientific fraud and scientific misconduct have gone through review and were detected only after other groups tried and failed to replicate the published results. An example is the case of Jan Hendrik Schön, in which a total of fifteen papers were accepted for publication in the top ranked journals Nature and Science following the usual peer review process. All fifteen were found to be fraudulent and were subsequently withdrawn. The fraud was eventually detected, not by peer review, but after publication when other groups tried and failed to reproduce the results of the paper.

A study published in the peer reviewed Lancet Journal associating long-term use of non-steroidal anti-inflammatory drugs (NSAIDs) with a lower risk of oral cancer, was shown to be "completely fabricated"^[23] , after which the journal published a retraction, and acknowledged that the study "contains fabricated data." ^[24]

Although it is often argued that fraud cannot be detected during peer review, the Journal of Cell Biology uses an image screening process that it claims could have identified the apparently manipulated figures published in Science by Woo-Suk Hwang.^[25]

Peer review and plagiarism

A few cases of plagiarism by historians have been widely publicized.^[26] A poll of 3,247 scientists funded by the U.S. National Institutes of Health found 0.3% admitted faking data, 1.4% admitted plagiarism, and 4.7% admitted to autoplagiarism.^[27] Autoplagiarism involves an author republishing the same material or data without citing their earlier work. An author often uses autoplagiarism to pad their list of publications. Sometimes reviewers detect cases of likely plagiarism and bring them to the attention of the editor. Reviewers generally lack access to raw data, but do see the full text of the manuscript. Thus, they are in a better position to detect plagiarism or autoplagiarism of prose than fraudulent data.

Although it is more common than plagiarism, journals and employers often do not punish authors for autoplagiarism. Autoplagiarism is against the rules of most peer-reviewed journals, which usually require that only unpublished material be submitted.

Abuse of inside information by reviewers

A related form of professional misconduct that is sometimes reported is a reviewer using the not-yet-published information from a manuscript or grant application for personal or professional gain. The frequency with which this happens is of course unknown, but the United States Office of Research Integrity has sanctioned reviewers who have been caught exploiting knowledge they gained as reviewers. A possible defence (for authors) against this form of misconduct on the part of reviewers is to pre-publish their work in the form of a preprint or technical report on a public system such as arXiv. The preprint can later be used to establish priority.

Dynamic and open peer review

It has been suggested that traditional anonymous peer review lacks accountability, can lead to abuse by reviewers, and may be biased and inconsistent,^[28] alongside other flaws.^[29]^[30] In response to these criticisms, other systems of peer review with various degrees of "openess" have been suggested.

In 1996, the Journal of Interactive Media in Education^[31] launched using open peer review.^[32] Reviewers' names are made public and they are therefore accountable for their review, but they also have their contribution acknowledged. Authors have the right of reply, and other researchers have the chance to comment prior to publication. In 1999, the open access journal Journal of Medical Internet Research^[33] was launched, which from its inception decided to publish the names of the reviewers at the bottom of each published article. Also in 1999, the British Medical Journal^[34] moved to an open peer review system, revealing reviewers' identities to the authors (but not the readers),^[35] and in 2000, the medical journals in the open access,^[36] published by BioMed Central, launched using open peer review. As with the BMJ, the reviewers' names are included on the peer review reports. In addition, if the article is published the reports are made available online as part of the 'pre-publication history'.

The evidence of the effect of open peer review upon the quality of reviews, the tone and the time spent on reviewing is mixed, although it does seem that under open peer review, more of those who are invited to review decline to do so.^[43]^[44]

In June 2006, the high impact journal Nature launched an experiment in parallel open peer review — some articles that had been submitted to the regular anonymous process were also available online for open, identified public comment.^[45] The results were less than encouraging — only 5% of authors agreed to participate in the experiment, and only 54% of those articles received comments.^[46]^[47] The editors have suggested that researchers may have been too busy to take part and were reluctant to make their names public. The knowledge that articles were simultaneously being subjected to anonymous peer review may also have affected the uptake.

In 2006, a group of UK academics launched the online journal Philica, which tries to redress many of the problems of traditional peer review. Unlike in a normal journal, all articles submitted to Philica are published immediately and the review process takes place afterwards. Reviews are still anonymous, but instead of reviewers being chosen by an editor, any researcher who wishes to review an article can do so. Reviews are displayed at the end of each article, and so are used to give the reader criticism or guidance about the work, rather than to decide whether it is published or not. This means that reviewers cannot suppress ideas if they disagree with them. Readers use reviews to guide what they read, and particularly popular or unpopular work is easy to identify.

Another approach that is similar in spirit to Philica is that of a dynamical peer review site, Naboj.^[48] Unlike Philica, Naboj is not a full-fledged online journal, but rather it provides an opportunity for users to write peer reviews of preprints at arXiv.org. The review system is modeled on Amazon and users have an opportunity to evaluate the reviews as well as the articles. That way, with a sufficient number of users and reviewers, there should be a convergence towards a higher quality review process.

In February 2006, the journal Biology Direct^[49] was launched by Eugene Koonin, Laura Landweber, and David Lipman, providing another alternative to the traditional model of peer review. If authors can find three members of the Editorial Board who will each return a report or will themselves solicit an external review, then the article will be published. As with Philica, reviewers cannot suppress publication, but in contrast to Philica, no reviews are anonymous and no article is published without being reviewed. Authors have the opportunity to withdraw their article, to revise it in response to the reviews, or to publish it without revision. If the authors proceed with publication of their article despite critical comments, readers can clearly see any negative comments along with the names of the reviewers.^[50]

An extension of peer review beyond the date of publication is Open Peer Commentary, whereby expert commentaries are solicited on published articles, and the authors are encouraged to respond. The BMJ's Rapid Responses^[51] allow ongoing debate and criticism following publication.^[52] By 2005, the editors found it necessary to more rigorously enforce the criteria for acceptance of Rapid Responses, to weed out the "bores".^[53]

Peer review of policy

The technique of peer review is also used to improve government policy. In particular, the European Union uses it as a tool in the 'Open Method of Co-ordination' of policies in the fields of employment and social inclusion.

A programme of peer reviews in active labour market policy^[54] started in 1999, and was followed in 2004 by one in social inclusion.^[55] Each programme sponsors about eight peer review meetings in each year, in which a 'host country' lays a given policy or initiative open to examination by half a dozen other countries and relevant European-level NGOs. These usually meet over two days and include visits to local sites where the policy can be seen in operation. The meeting is preceded by the compilation of an expert report on which participating 'peer countries' submit comments. The results are published on the web.

U.S. government peer review policies

History of peer review

A practice similar to a peer review process is found in the Ethics of the Physician written by Ishaq bin Ali al-Rahwi (854–931) of al-Raha, Syria. His work, as well as later Arabic medical manuals, state that a visiting physician must always make duplicate notes of a patient's condition on every visit. When the patient was cured or had died, the notes of the physician were examined by a local medical council of other physicians, who would review the practising physician's notes to decide whether his/her performance have met the required standards of medical care. If their reviews were negative, the practicing physician could face a lawsuit from a maltreated patient.^[58]

Peer review has been a touchstone of modern scientific method only since the middle of the 20th century, the only exception being medicine. Before then, its application was lax in other scientific fields. For example, Albert Einstein's revolutionary "Annus Mirabilis" papers in the 1905 issue of Annalen der Physik were not peer-reviewed by anyone other than the journal's editor in chief, Max Planck (the father of quantum theory), and its co-editor, Wilhelm Wien. Although clearly peers (both won Nobel prizes in physics), a formal panel of reviewers was not sought, as is done for many scientific journals today. Established authors and editors were given more latitude in their journalistic discretion, back then. In a recent editorial in Nature, it was stated that "in journals in those days, the burden of proof was generally on the opponents rather than the proponents of new ideas."^[59]

Peer review of software development

Contents